Speaker 1Hello, everyone, and welcome back to the Travel Trends podcast and a special spotlight episode. Today, we're going to look at the world of AI in travel, but from a very unique perspective. We're going to bring in Manuel Hilti, the CEO and co-founder of Nazaza, the AI travel packaging platform behind TripBuilder. You can find out more information about them as we're having this conversation at nazaza.com. That's N-E-Z, I'm Canadian, of course, or Z for those Americans, N-E-Z-A-S-A.com. And what's really fascinating about Manuel, he's based in Switzerland, but his background, this guy has a PhD in computer science. And before turning his attention to travel technology, which I'm grateful he's now in our industry, I've got a chance to know him over the last couple of years, and we've been together at some of the conferences. And the more I get to know about Manuel and Nazaza and what they're building and some of the exciting launches, but this all goes back to 2012. That's when TripBuilder was introduced, and they work with some of the biggest names in the travel industry. One example of that is TUI. Many people know TUI is one of the largest global travel companies, and they're a major partner with Nazaza. And what Nazaza really does is they put AI to work behind the scenes with their tools like planner, co-pilots. But today. And the reason we're doing this spotlight episode is that Manuel is here to introduce what they've just launched, TUI Workbench, a new AI platform built specifically for travel operations. We're going to get into what that means, how it's different from bolting a chatbot onto existing software, and why Manuel believes travel companies should start thinking about AI less as a tool and more as a digital workforce. I couldn't agree more. He's actually going to be part of our Travel Trends Beyond AI Summit, October 27th. 28th and 29th. He's presenting in the afternoon of October 28th. So if you're listening to this before then, come and join us. It's free to attend. If you're listening to this after the fact, I've got great news for you. You can instantly go and watch that session on YouTube and be able to see beyond the spotlight all the really cool things that Manuel and the team at Nazaza are doing. So on that note, Manuel, welcome to Travel Trends. Great to have you with us. Hello and thank you. I'm very, very glad to be here. Fantastic. Tell everyone where you're based. I think that would be. I mentioned Switzerland, but many people have traveled to Switzerland. Tell everyone where you're based, where the team is.
Speaker 2Yeah, our office is actually in Zurich. I'm personally living a little bit east in the northeast of the country, but we're Zurich-based. And we. Like in Switzerland, we have people from all over the country who a couple of days per week take the train into the city and meet up there. Most of our people are actually based in Portugal. So our biggest office is in Lisbon. And this is where almost two-thirds of the company are.
Speaker 1Fantastic. And obviously, you're from Switzerland and you did. I mentioned about your PhD in computer science. And the reason I wanted to emphasize that is because there's very few people in the travel space that have that same kind of pedigree and technology background. Because obviously, people with that expertise go into different industries. And I know when you were starting your career, you were in California doing your internship at Hewlett-Packard way back in the day when HP was a big deal. You've worked for banks. And so you've got a really good. Understanding of technology. But you decided to apply that to travel. And I'd love to understand that first. What attracted you to the travel technology space?
Speaker 2Well, I mean, part of it was a little bit serendipity, you know. I actually, in a couple of years before we started in Azaza, I was actually doing consulting. So we had a little boutique consultancy with two other business partners where we helped companies set up projects at the business IT interface and tried to avoid. A lot of mistakes that people usually do there. But I always knew that I didn't want to be a consultant for too long. I wanted to do my own thing. I wanted to have my own baby. And I was kind of constantly looking around for, you know, for an opportunity to build a product. Now, another thing is that I did a lot in finance, like fintech and things like that. And when you work for financial companies in tech and then. You know, you're at a wedding, at a party, you meet people and they ask you what you do and then explain to them what you do. And then it's kind of not so exciting. And so travel was always. Travel is something that everybody can relate to. You know, everybody travels. Everybody loves to travel. Everybody's eyes light up when you travel. And so when we stumbled over the travel industry and got to know people from the travel industry, then it kind of had a certain attraction to say, OK, let's try to make travel better. Of course. We might not have done it if we had known at that time how complex travel is. On the outside, it looks very simple, you know. Like you think, OK, it should be easy to do it better. It should be easy to fix. And then you enter the travel industry and then you figure out, oh, it's actually way more complicated than I ever thought. But otherwise, maybe you would never get started, which is fine.
Speaker 1Well, that's just it. I mean, if I look at a common thread between the guests we've had on our podcast and the entrepreneurs in travel technology, that I've known over the years and come to meet through the podcast, there is certainly a through line, which is that that complexity is actually a source of attraction and maybe even actually what's kept you engaged in the industry is because it is messy. It is built on a lot of outdated, antiquated technology and you doing computer science and, you know, wrapping your head around the fact that we had these GDS systems, global distribution systems. The travel industry was built pre-internet on these platforms like Sabre and a lot of that. So plumbing is now it's often referred to is what is being rebuilt now in the age of AI. So it's an incredibly exciting time to be in this industry. It was obviously when you started, even I would say it's that much more exciting now. But what I'd be keen to understand is that as you transition from having a consultancy business to actually building out Nazaz, and obviously it's primarily known as a trip builder, the whole idea of a connected trip. And I want to explore that a bit because that terminology was something that Booking.com was using for many years, as you know. And it's like, when is this connected trip? Ever going to happen? You guys have really built the connected trip experience for your clients. So that terminology really makes sense in what the problem you were solving. But tell us, what was the original problem you set out to solve in travel technology? And give us a little bit of how that's evolved over time.
Speaker 2Yeah. And fundamentally, the problem was very similar in the beginning. So what we wanted to do is make the planning and booking of end-to-end trips easier. So that was what we set out to do. However, like pretty much every travel startup at that time, we wanted to go B2C. So we actually built our own website. We sold trips to end-to-end, high-value, one-to-three-week trips to Asia to customers directly. So we were a travel agent with building their own technology. And then we did that for a couple of years until we realized it was not the right business model for us. So our DNA was. It was the one of a tech team and not a consumer marketing team. And then we said, okay, like what we've built here and the technical problem we tried to solve here is actually something that the travel industry could use as well. And then we pivoted to a B2B technology model around 2016. But we still, I mean, in that phase, we made a lot of experiences of what it actually means to sell travel. And I think we still somehow profit from that experience today.
Speaker 1For sure. And I guess that's where, when we look at the core business, the idea of, you know, a packaging platform for the connected trip. And we look at the types of clients you work with, airlines, tour operators, and various enterprise. I mean, the core capability of the platform being, you know, frictionless planning, seamless booking. Obviously, those are the things that so many travel companies are looking for. Because many, many or most, I would say, travel companies are not technology companies. And this is something that I often talk about. When I'm on stage and when we're doing these types of conversations is that people who love travel and get into travel for all the right reasons they want to work in the industry, a few of them have technical backgrounds. So when you actually start to build up a travel business, you really do need strong technology today. And some of the most successful travel companies have a strong technology backbone or make sure from the beginning they find the right technology partners to work with. So as we're kind of discussing a little bit about the background, tell us about the customer, the customers that you started off with. Obviously, I mentioned TUI. Obviously, they weren't necessarily your first customer because obviously they are a high-profile one that you eventually brought into the business. But how did your platform evolve? What customers were you serving first? And then how did you kind of build it and scale it and expand it into different verticals?
Speaker 2Yeah, so our customers were tour operators in the European style in the beginning. And the first two customers were very small. We had a very small specialist tour operator and a very large German tour operator back then. I think it was the fourth biggest German tour operator. So we already started at both ends of the scale in terms of size. And then we actually continued working with these European style tour operators up until the pandemic. Mostly with medium size to large ones. So we very early had to learn how to scale. We had to integrate into existing landscapes where there are other systems that they use and we have to fit in. APIs to connect to them and all of these things. And actually, TUI came in at the end of the pandemic. So that was kind of phase two then of our development when we signed up TUI.
Speaker 1Well, it seems to me, having met some of your team, I've been continually impressed as I got a chance to meet more of your team members. Obviously, we've met multiple times and I was very keen to have the spotlight episode to help understand the background, Nazazo, where you've gotten to and where you're headed. Because I think one of the things that obviously is going to be really exciting for people to hear in this discussion, then also part of our summit, is how you've built in AI. Because it seems to me that you've kind of really supercharged your business with the integration of AI. For anyone who's looking at TripBuilder on your website today and looking at all the AI functionality you've built into the platform, obviously that didn't exist three years ago or even when you were just post-pandemic. So much of this has happened in the last couple of years. Tell us a little bit about TripBuilder. Because since that's the core platform, and I've mentioned it, but I'm sure our listeners would be keen to know what TripBuilder is, what the functionality offers. Obviously, I've kind of mentioned the clients it's for, but give us, Manuel, an overview of TripBuilder. Yeah.
Speaker 2So TripBuilder, in essence, is a flexible packaging and reservation platform. So you can use TripBuilder to put together any type of package, itinerary trip, and then book it and manage it post-booking. And this just with a very high degree of automation. We started out on a more complex, tailor-made side where you have those like 10-stop self-drive itineraries and the complex, I don't know, trips through Asia and all of that. But we have actually since like covered the whole packaging space. We can now do like the classical dynamic packaging, which is a flight in a hotel. We can do single-component hotel only. But you can always. The main philosophy is that it should allow our customers to sell their clients what they want, when they want it. So it shouldn't be constrained by it's a platform that can sell this type of trip or that type of trip. It should be able to sell any type of trip through any channel at any point in time. And this flexibility is really what we care about. So in terms of customers, by the way, because we also now support simpler types of packages and more dynamic types of direct-to-consumer workflows, one new segment for us is, for example, now also airline holiday packaging. So that's another vertical we're currently quite busy in, in addition to the traditional tour operator in the agent space.
Speaker 1And one thing I think is worth calling out for our listeners, too, is that as you've built out this platform, you were solving for tailor-made itineraries that were kind of online or offline. And this whole idea of. Empowering travel agents as well as tour operators to have more of an omni-channel platform so that they can take their products and their expertise and kind of work online, offline, because obviously it is such an important part of our business. I wanted to highlight that because so much of our business is still human touch and travel advisors working with their clients and bookings occurring offline. And so although all the products are bookable through the platform, the whole idea is creating a process where you can. Hyper-personalize and package. Is that a fair description? What else would you add to that that would be interesting for our listeners to know?
Speaker 2Yeah, it's a fair description. And the channels can also work hand-in-hand. So there are also hybrid models where a customer, like an end customer, could go to one of our customers' websites, start planning an itinerary at some point, say, hey, I need advice. Can an agent please help me? The agent can access the same itinerary, can modify it, can play it back to the customer. So those channels are not. It's mutually exclusive, but you can actually work across channels as well.
Speaker 1Yeah. The reason I wanted to highlight that is because so many travel agents out there, and obviously that is a big contingent that listens to our Travel Trends podcast. I'm thankful and grateful for that because that's how I started in the industry on a travel technology side, implementing tools for corporate travel platforms, working with travel agents to better understand their role and what we could automate and how we could offset their workflow. And so I find that really interesting because the first thing that all travel advisors need to do is. Build out itineraries for their clients. And your core competency with using AI now, obviously, to create personalized itineraries at scale, many travel advisors are just starting to embrace some of the large language models to figure out how to do this themselves or how they can better create itineraries because it's massively time-consuming. So that's one core part of the functionality that I just wanted to highlight because everyone is spending a lot of time on trip planning for their guests, and this is one of the benefits of the Nazazo platform. Before we move on to this big launch, I'd love to kind of set the stage because it almost seems like, from my perspective and getting prepared for our conversation today, I was going through your website, and although we've met and spoken to the team, I was learning a lot more about the company and obviously some of the exciting launches. And for me, Manuel, it seems like there's kind of been three phases to the business. There was kind of like life before AI. And then there's been the last few years. There's been the last few years where you've really started building AI, and that's why I'd say really scaling the business. And then this next phase where it's like, you know, you really are transitioning the company with the launch of Tai to become, you know, an AI-native or fully AI-led business. But tell everyone, if you wouldn't mind, just why you're making that leap and how that all came about between what you had built with TripBuilder and all the AI integration to this major launch of Tai, and then what's that going to mean going forward? So, I'd love to hear from you. So, I'd love to get into that next, but tell us how you, you know, what was happening on your end with the team and your vision for the business that you were kind of like, this is the next quantum leap for us.
Speaker 2Yeah, I think one very important and retrospectively also kind of a little bit lucky decision, because, I mean, we didn't know what the implication of it would be, is that about four years ago, I think it was actually even more technical than I am. So, he knows more about all of these things than I do, even though I do have a technical background as well. We decided that he should fully focus on building the next generation data infrastructure for Nisasa. And this is already when AI was very much on the horizon. So, we said, okay, let's first build a very good data layer, which will be, will be the foundation to build a lot of, a lot of AI on top. And basically for the last four years, all he has done is work with data and AI. And we have very early started also internally, like a core group of people started to build lots of internal tools, also in our engineering, basically AI bots that help us much better deal with support cases, automating our software development lifecycle. So, we started these things very, very early. In the course of that, we then also started to build first AI functionality into TripBuilder. And the first thing we did, we called it Planner Copilot, is a functionality to basically turn any type of itinerary description into a bookable itinerary. So, it's less about creating the recommendations because other people can do that way better than we do. We don't know the end customer, but we can take all types of input, be it in text, in image, in video content, whatever, like this. This is the itinerary that I would want, and we turn it into a bookable itinerary. We can also use that to make agents way more efficient because when an agent has a customer on the phone, they type their notes about a conversation into Planner Copilot, and then they get an 80, 90 percent of the itinerary they actually want to sell to the customer, drastically basically reducing the time and work they have to put into this. So, this is kind of the history. So, at the end, at the end of last year, we then said, OK, we have this, we have, we basically AI-enhanced trip builder. We have all of that know-how at building, like, very, very advanced internal AI tools. And what's next? And then we said, OK, the next, the next thing is to build an AI-native product that solves a very hard problem. And we very, like, we did a lot of thinking about what will be commoditized with AI, what will be very easy to build with AI, and which are the new ones. New types of problems that we couldn't solve in the past and we can actually solve now. And everything we did since then in terms of the new product is an attempt to say, OK, we have this new tool and let's now use it to solve problems that before we just were not able to solve.
Speaker 1Yeah. And the reason I wanted to ask about that is because from my perspective, seeing Nazaza and looking at the next wave of travel technology as travel companies have embraced agentic AI. Over the last year, and obviously, agentic AI has been the big talk over the past year. You know, we're three, almost four years into this, you know, as my friend Gilad Berenstein would refer to this next epoch of travel technology, where we started off with the large language models that could create content and now they can act on our behalf and the whole idea of agentic. But clearly, this requires a complete rethink of how you structure your business, how you structure your teams, what everyone's roles are and the more. successful companies I'm seeing in the industry, um, tour radar being an example of that. Captio, there's a number of companies that are railbookers and these are people that will be at our summit as well as Manuel. They're looking at their business and understanding how profoundly this is going to change all aspects of how they operate and encouraging their team members to embrace it and to take advantage of this technology. But really, it does require, so it's interesting to get that background and perspective on your friend, colleague, business partner that obviously was kind of shaping up the next wave that now has been announced as Thai Workbench. So tell us about Thai Workbench. Obviously, you know, why did you decide now is the time to launch AI Workbench? I mean, the phrase itself is very intriguing, but I just wanted to say one last thing because I think this will hopefully be helpful for our listeners too, is that, you know, we have already seen a first wave of travel AI companies that have kind of been a wrap-around. And I think that's a really good way to start her to LLMs. And that business model has been very short-lived because the LLMs are then quickly replacing those platforms, just like Google decided to add weather and sports results. It was like an interesting business for a short period of time. But that window was closed very quickly with the rapid progress of ChatGPT and Claude. And so this is where what you're proposing now and what you've built is a much more profound shift. So I just wanted to set, you know, the stage for that. The idea that you have an AI Workbench is exactly how companies should be thinking about their business today. So, yeah, tell us about Thai Workbench and what it means and ultimately why all of our listeners should be intrigued by this.
Speaker 2Yeah. I mean, what our thinking process was a little bit, I told you, we were looking for kind of the underserved heart problems in travel. And one thing that was pretty clear from our experience working, with our customers is that as soon as you sell a complete trip, a package, an itinerary, whatever, the operational overhead and the operational risk that you assume as the merchant of record of this trip is what's eating away a big part of your margin. So the operator business is not going to go away because people in like the search behavior of people is more in complete trips and they, they ask an AI platform about what should I do? They get a recommendation for a complete trip. So there is a natural desire to actually buy a complete trip from an agent or operator. But this entity then has to live with all the consequences. And we saw a pretty extreme situation again with the war in the Middle East breaking out and a lot of our customers losing a lot of money with all the operational overhead that comes out. There is also all of that small little frame of mind that we have, you know, we have to live with all the consequences. And we see that friction every day from, you know, problems in flight schedules or from natural disasters or strikes or border problems and all of these things that always create this little friction. And this operational overhead is just huge. And it has been completely overlooked by technology in, in our opinion. So we then said, okay, we want to build something that reduces the cost of the flight and the cost of being an operator of a package or itinerary. And basically, this led to two things. One of them is, we'll talk about that later, is kind of the, the flagship use case that we want to solve. It's making sure that we can reduce the risk, manage and reduce the total risk of an itinerary operational risk over the whole lifetime. And when we started building that, we had to build a foundation to build this agent. Then this foundation became a product in its own, which is this AI workbench. And you can think of it like a little bit, a cloud cover specifically for travel, which knows travel, which is integrated with travel sources, which can operate a lot of those operational workflows, use cases, edge cases that always occur in travel businesses, which costs a lot of effort and money and slow everybody down. And this is what we want to give our customers a tool to automate all of that. And the reason why it's not an agent, but actually a workbench where they can create their own agencies. And one, one of the big learnings in this industry is that everybody has different workflows, different problems, different edge cases. And in the past, we always struggled with this customer asking, can you build this? Another customer asking, can you build that? And we always say we, we cannot implement each of those custom use cases. And now they can do it themselves.
Speaker 1Well, and I think this is the profound shift that from introducing itineraries. And if you think about all those travel advisors that are creating itineraries, it's fantastic to create a custom itinerary. But now you need to actually coordinate the suppliers, you need to extend, obviously, considerable amount of customer service time. So it's beyond the content production. And that's what really intrigued me, seeing what you have, how the business has evolved. And then now, the idea of Thai workbench, I see it as a real game changer for the business. I like the branding, too. Obviously, I'm a, I'm a marketer at heart as well. So Thai obviously has T for travel and T for travel agent and AI built right into it. And so the idea of Thai workbench. But one of the things obviously, is that given your background in travel, one of the challenges we often have is that a lot of companies don't necessarily understand the industry when they get into it. And so and if companies are working with generalist technology providers, they don't necessarily understand the industry when they get into it. That don't really understand travel, they're learning travel at the same time as they're, you know, building something for you. And so the fact that you guys have been around since 2012, and you've had these, you know, these series of evolutions, you know, you've built Thai workbench to be purpose built for the travel industry, because it, you know, it has to solve these complex issues, like multi leg itineraries, and these different supplier dependencies that exist, and also the different booking life cycles of a customer. So like, there's so many different aspects to what a travel industry is. travel advisor does, which is why their roles are as relevant as ever now still in the age of AI. Because it is completely incredibly frustrating often to book these types of trips on your own and piece everything together. So that that combination of technology and human expertise. But tell us, because one of the things that stood out to me too, is I know you are Thai has been built as a LLM agnostic. So it's not tied to a specific LLM speaks a little bit more about that. Because I'm intrigued, because if travel advisors are listening to this, and then they're, you know, I find from my keynotes and my stage presentation, and people come up to me afterwards, they've moved past the beginner phase, most people are low intermediate on the scale of their knowledge and use of AI, everyone's using it. And they're using it. So it's, you know, we're beyond the beginner phase, we're moving into a more intermediate phase. And you see some more advanced use cases of it around the industry. So they're certainly familiar with Claude co work. But tell us more about this. So they can also understand, why would I use this versus some of these AI tools that I've started to embrace?
Speaker 2Yeah, so Thai does not just work in the context of one LLM, it sees LLMs as tools to invoke. And the thing is, I mean, from kind of an AI development perspective, you know, you had distinct phases, how, how this evolved. So first, it was all about prompt engineering, like, you had to learn to write the right prompt. And then it was about context engineering, you had to give the LLM, the right information context to actually solve the problem. And we're now, like kind of anthropic, then launched the, the area of harness engineering, which means you build this tooling layer that in the end calls LLMs in various shapes and forms, or orchestrates the whole process, and brings the problem down into sub problems that maybe not each of the sub problems you send to the same model. And what we can do now is, for example, things like, if you have a problem that typically you would send to a very, like high powered reasoning model, like an Opus, or the top Gemini or the top open AI model, which would maybe even give you a good answer in the end, but it takes a long time and costs a lot of money, right? And if you break the problem down into small problems, then you can send in each of those small problems to the cheapest, fastest model out there. And you, and then you pierce it back together again. And then it's not only cheaper, but it's actually way faster than if you just throw the whole problem at an LLM. And this is one of the things that we do on this so called harness layer. We also give it a lot of intelligence about figuring out which piece of the problem you actually have to send to an LLM for reasoning and which piece of the problem we can you solve without doing that, give it the right context or right data. So there is there is a lot of these things. And of course, security and compliance and privacy and multi tenancy, and all of these things you need to solve as well, right?
Speaker 1Yeah, and obviously, that's one of the things I found fascinating was the ability for to create AI agents that will just understand plain English. Obviously, the no code is incredibly important. The one thing, when you look at the repetitive tasks, when you are analyzing the workflows, of the tour operators and and travel advisors um tell us what would be some of the repetitive tasks that would otherwise consume i mentioned obviously putting an itinerary together but that's only one part of the what other repetitive tasks that travel advisors are currently doing right now um does this solve for
Speaker 2the funny thing is that we kind of discover more and more of them um from giving this tool to our customers so it's actually it's actually quite fascinating how you you give it to them and then you run sessions with them and then they come up with a lot of use cases that we were not even aware of that like eat up a lot of a lot of manual time um so so it's it's kind of you know very simple things like pre-departure checks like x days before departure of of a client you need to go in look at the itinerary is this done is this done is this in order all of these things you know and sending information about an upcoming itinerary to somebody be it the driver who has to pick the people up in the morning or be it somebody in the destination who needs to know something sending reports to to dmos every one to three months about how many packs you send there um loading rate sheets is like a terrible thing right we cannot do that yeah well so it's like the thing is it's it's this death of a thousand cuts you know i mean there is one very big thing that we want to solve we can which we can talk about afterwards with our our flagship tie guard agent which we're about to launch but the rest is really the just those many small sometimes very custom processes that so far nobody automated for them because everybody has different processes and each of them is kind of too small to be built dedicated software for but if you now can help yourself and put up agents who automate all of these small little things you instantly gain a lot of time like a lot of resources back and you're also much faster serving your clients
Speaker 1and that's where i've been and i was interested to ask you that because obviously you're starting to see the wow factor from clients that are embracing this and some of the practical ways that they're embracing this technology and realizing they can use it and they're embracing it and they're embracing it and they're embracing it utilize it one thing also is an important to note too is that um and this always comes up when we're talking about ai as data and obviously they have the one thing that you had already built into the platform was the idea of having access to data and then being able to leverage the data that you had access to to be able to create all these workflows um before we get into tie guard i'd love if you wouldn't mind speaking to that a little bit because one of the challenges obviously when someone's using an lm outside of their existing systems that they have their crm platforms um you know exporting importing data how they're sharing their data especially when it's sensitive client details um you have to be really mindful um of how you're using you know lms today i think most people are aware of that not that you know there always is that concern that it's going to get trained off your data and um my good friend christian watts who's a you know a pretty well qualified ai expert in his own right would always highlight that it doesn't really care about some of the individual like it's not trying to steal your data but the point of it is you know there is that risk for sure um they have to be concerned about um so but in terms of being able to get the full benefit of the platform you really need to have data integrated so i just i'd love for you to speak to that because i think that's that's one of the unlocks that i see that you already have with being able then to layer this in and rebuild your platform around you know being an ai native business but you've already got the the data layer yeah i
Speaker 2mean about your first point i mean that's that's another benefit of being llm agnostic right um like first of all like our standard deployment model here is that we use models hosted in amazon bedrock and not the like the primary instances of the the llm providers so all of those models are purely hosted for inference and not for training so like that's kind of one thing solved but you can also choose which model to use you can also take an open source model you can um you can actually pick when you create an agent which ones you you want to use for it um if you don't want the harness layer to to optimize and always pick the right one for you you could in theory also host them somewhere else and point it there and say this this is the model deployment i want to use so so this abstraction actually also helps with the very sensitive clients and that's
Speaker 1exactly why i was um interested to ask this because obviously this idea of connecting to booking data from day one and that being part of the packaging platform um giving customers access to their booking data right out of the box like customers that embrace this so tell us a little bit more about how you guys actually do that
Speaker 2yeah i mean there are various ways of of getting data in there one of them is by connecting it with a system so tie is natively integrated with trip builder but it's actually not married to trip builder you could use tie with your own booking system connect to his apis getting like access to booking data do something with it and we have a very flexible no code connector framework where customers can actually do those integrations themselves without us having um to write code for them you can also make your own internal data available to it you can actually um load your own company knowledge into a vector database make it available to tie so that it makes the right recommendations at the right point in time also very useful for planner co-pilot where for example you can you can load your distilled agent knowledge into the tool always protected no other customers will ever get access to it and then you will make recommendations that actually follow your internal guidelines of of what you want to recommend to your clients and then it's also very easy to connect to additional data sources and we've connected to a lot of them we have and we already have flight data from oeg in there we have data from providers that um that map destination risk we have traffic data we have like a lot of a lot of data sources we have in there and we can actually leverage them in in all of those agencies and
Speaker 1then you mentioned tie guard and just as a bit of a segue to that because what i've seen and understand from the work you've been doing is that when i mentioned customer service there's obviously um a significant amount of time that's invested of uh the back and forth like with resolving booking queries and i've always seen this is you know where travel advisors and tour operators get bogged down and um even on trip you know i sit on the board of a tour operator in columbia one of the challenges we just talked about at their last board meeting um what's the most important thing that you can do to get the people who are doing uh self-guided trips are still continually asking for restaurant restaurant recommendations and they want like a more of a concierge service and so it doesn't it doesn't finish when they get on the trip and it's it's still you know uh customer service is paramount and often when i talk about this um at conferences i'm always very mindful of the fact that you do not want to train your customers to use the chat bot instead of their travel advisor because you are essentially you know pushing them away you want the travel advisor to be empowered by this technology but that's why i'm keen to ask you this when if you're starting to see some of the early adopters uh being more efficient with being able to resolve booking queries um tell us a little bit more about how that is working today and then how that relates to where you're headed next with tie
Speaker 2guard yeah okay two things first of all i'm a strong believer that like automating a the nitty-gritty tasks internally actually allows people to talk more to people so so we have the same in our company we have this this um like great internal tool we call sherlock which is a little bit um a cousin of tie and which can um analyze any support case and tell you is it the user problem is it the supplier problem is it a problem in our platform if it's a problem in our platform it will even pinpoint the lines of code where the problem occurs so it's a huge time saver because of that we we didn't um like say okay now we don't need support agents anymore what we said is okay this team needs to speak more to customers right because they don't have to spend hours investigating they can finally spend more time with customers um which is actually very very important very very valuable and the travel advisor can use tie to to serve up any information about an itinerary they can ask it tell me what's happening in this part of the itinerary they can ask how much money do i make out of this um why has this changed who has changed it when did the discount come on to the itinerary all of these things are very easy to answer now so they that they can spend more time with their customers and what we're doing now is we're taking this one level further from a single itinerary context to your whole portfolio of travelers, right? basically what we also want to do is first of all prevent friction but then also make it easier to deal with friction and this this leads us to our flagship agent on thai which we call thai guard which analyzes monitors and mitigates risk in connection with the trip so in the end it's kind of one of the core pieces that should finally deliver this fully automated connected trip because because it fixes itself right and it starts with analyzing an itinerary for potential issues it can tell you things like okay the three hours between landing in bangkok and taking the ferry out to the islands are not enough because this flight that you want to book here is on average one hour delayed and then you're landing at the time of heavy traffic you need half an hour more to get out to the ferry terminal so you might want to book a different flight or a different ferry it can it can you can do a lot of things and you can do a lot of things and you can do a lot of things highlight a lot of other problems in connection with an itinerary before it's even booked and when it's booked it will continue to monitor all of your trips as an operator or an agent and for any new source of friction coming up will instantly flag when there is a wildfire when there is a new war when there is a strike announced when there is a change in flight schedules a hotel burned down it will always serve everything up that could be affected by this event it will tell you how much it is affected and later um this is um this will be in a later iteration of the product it will also propose the mitigation for it and at some point you will be able to tell it okay in the future you can action those mitigations autonomously and this should also help our customers to just deal with friction in a in a much better much more efficient way and in the end the traveler also profits because their potential problems will be detected as early as possible um which leaves more time to fix them before they become a real big problem you know and this is what we want to do yeah
Speaker 1i find this really interesting because obviously ai workbenches thai workbench is live now thai guard is uh going to be launching shortly so depending on when you're listening to this obviously you can go to nazaza.com to learn exactly what's live currently the part i was going to ask you about on thai guard was exactly what you were mentioning about where it fits into so you've got thai workbench you've got thai guard my initial thought man well was that thai guard becomes um once you actually move from planning into traveling but you just mentioned because i know you have an itinerary checker so that um you know the system obviously can um recognize um issues before they come up will thai guard then be just built right into the both will be happening from the very beginning that you'll be working um on building out itineraries for uh for guests using thai workbench and thai guard will be there from the beginning or does it come in at an at the next stage of the planning or traveling process because you mentioned about it coming in to check the itinerary before you're traveling and to make sure because real-time updates is a real is a real thing with traveling making sure flights and it's like travel is not it's not so you get home and you will be like you're looking through the photos and like but a lot a lot of stuff there's a lot of challenges that can come up with travel and people i've it can be for some reason people don't travel i think it's too high risk to travel but obviously we love traveling and you have to just accept some of those risks and sometimes you have things bumps that happen on your trip um but sometimes those are the best stories when you get home but obviously from a travel advisor you're trying to make sure that trip goes as smoothly as possible so where does thai guard come in to play yeah so thai
Speaker 2guard is a little bit the the little guardian angel that goes with an itinerary through the itinerary so it's a little bit of a little bit of a little bit of a throughout the whole life cycle right and it coming can come into play in planning um it will be natively integrated with trip builder already in the planner so you can fire it up and tell it tell me ask it tell me what's wrong what's potentially wrong with my itinerary it will flag it during the checkout and then it will continue to monitor and to you know take care of your itini throughout the life cycle but it won't only work with trip builder you can also create it with another booking platform and then for example when you enter the checkout you can um you can uh trigger thai with your itinerary you can ask it what's wrong it will give you an analysis you can show it to the agent or even like a simplified form to the end customer and do something with it and it can still actually continue monitoring your itineraries also in connection with other booking platforms so it's not it's not um exclusively tied to trip builder the only thing that right now we don't yet know how we should do it with other booking platforms is actually automatically fixing things because that's an area where trip builder is super strong and gives a lot of um a lot of tooling and functionality to actually change itineraries after booking and this is something that depending on the booking system you use you might have more um options there or less but the rest um is kind of something that you can attach to an itinerary and then you can actually do it with another booking platform itinerary very early in the life cycle and then you will stay with it until the end interesting
Speaker 1well i guess what i'm curious to know then as you build out thai workbench thai guard and you look at the uh utilization of those with your existing clients um whether that will lead to expansion into other verticals you know i mentioned tour operators and airlines kind of being a core focus for nazaz that today is the intention with this technology to deepen those categories obviously there's still a lot of time to do that but i think it's going to be a lot of time to do that and it's going to be a huge opportunity for growth just with an airline you could literally just choose one of those markets just tour operators just airlines they're massive industries unto themselves um what i'd be interested to know is whether or not when you start looking at that technology and where you're headed with does this enable nazaza to expand into other verticals is that part of the future plan uh yes it's
Speaker 2definitely something we're heavily considering i mean if we look at thai then it can basically do two things one of them is it can make trip builder and it can build a much better but it can also allow us to provide a product to customers who might for whatever reason not use trip builder which is kind of a heavier product in this trip builder goes very deep into a customer's core business processes so it's a little bit it takes a while to adopt because they really have to do a transformation there whereas thai you can and both the workbench and the thai guard you can basically just bolt onto your existing system landscape and with thai guard we absolutely have to do a transformation there and we have to do a transformation and we have to do a transformation there and we have to do a transformation there believe that there are a few more segments where this could be very interesting um and we would be very open to speak to a lot of people from those segments to validate the proposition and um i mean that's definitely all types of agents host agencies um and and these types of companies um it's also tmcs who also have like um a big interest um to tell their corporate clients how risky um both operationally and from other perspectives the trips are that their their employees book um and and also otas who sell more than just a flight or a hotel so we see more segments that could actually profit from this um risk management aspect of thai guard and but but this is really just starting so so this is something that we're going to be exploring the next few months
Speaker 1well let's talk about the future and keep on that um topic for a moment because um obviously things continue to change very rapidly and i've called out a couple of elements that stood out to me when i was looking at the platform and getting a better understanding of where you are and where you're headed um because uh no one wants to build a platform that will be obsolete six months from now because these ai models are evolving weekly and so um you know we talked about data and how they be being agnostic in terms of llm um when you look forward um you know in 27 28 and where nazaza is headed um clearly obviously expanding into other verticals what what stands out to you as far as the the bigger picture about how ai you know when we first started our ai summit the first year it was about how ai was going to impact travel last year was about how ai is transforming travel and as you know this year's theme is beyond ai and the theme being you know ai is just now being built in it's to the uh the tech stack but there's also a need to make sure that the the um the travel experience remains innately human that is saying it's part of the human experience that's desire to travel and so i also want to be mindful of where ai plays a incredibly kind of vital role in these in travel businesses but also where the human touch is important and so i'd love to kind of get your view on that as far as you know where you're headed so i'm i'm kind of asking you a kind of a broad open question here because i'm sure these are the things you guys talk about as a team human expertise automation where this industry is headed and i'd just love for you to kind of share some of your your views on that as you look towards the future of nazazo
Speaker 2yeah so i mean in the end i personally believe that travel will always be about human connections on a certain level right and and they won't go away i mean the question is a little bit depending on the type of customer they will want less or more travel and i think that's a really um exposure to human connections and so so there won't be a a one-size-fits-all solution there will always be people who want to talk to a travel agent for various reasons but there have always been people who booked their trips without talking to a travel agent and i don't think that this picture is fundamentally going to change so we want to give everybody the tools to serve their customers better whether they want to provide them self-service or whether they want to make sure that their best agents get so much more empowered and so much more efficient that they can actually advise more customers that they can sell more and that they can just leverage their expertise in in many more ways and like our strategy is really to and this ties back into the omni channel we're not favoring one of the channels we want to make all of them better
Speaker 1yeah that's um the i mean the very concept of omni channel it's interesting when i when you look at what you just described it was the same problem you know 20 years ago in the dot-com era when all of a sudden this whole uh otas came up and it was like the end of travel advisors and the same thing was said again with the face of ai and i thought what you said there is important um that there were there actually was always the self-serve option and some people chose that and there was more self-serve there's never been more options and there's some people that embrace that and they book their trips on otas um and there's still the massive uh marketplace for people that work with travel advisors and want that human trust so they actually want humans involved in the decision making um the recommendations um and so it's interesting that you are well positioned to serve both because obviously i see in in terms of your strategy it's not one or the other it's both so it's like you're like your channel agnostic like they can go direct to a tour operator um and the tour operator uses your platform or they can work with a travel advisor and the travel advisor like it's so from your vantage point like both are opportunities for growth
Speaker 2for nazaza yeah and even cross channel so um some of our customers really also want to provide the option to either help yourself or talk to an advisor when you get stuck you know yeah when i was
Speaker 1preparing for our discussion there was a few things that obviously uh when i think about you know when i approach these conversations i'm always thinking about what our listeners want to ask and what like if they had the privilege of having this conversation with you manuel obviously you know people will reach out to you and we'll finish on that in a moment so that people want to partner with you or learn more about nazaza they can reach out um but while we're having this opportunity to speak i always want to make sure that people listening to this the kind of what questions would come to their mind and i i thought just given your expertise in the industry um i'd ask you a few more questions if you have any questions just about where the industry is where it's headed you know moving away from you know keeping your nazaza hat on of course um but also just looking at how ai is uh impacting our industry where you see the biggest opportunities for it so if we kind of take a step back and look at the the bigger picture um and think about how ai is changing the economics of running a tour operator that's one that i find really intriguing and i'd love to get your view on it because you work with tour operators and you're a tour operator and you're a tour operator and you're a tour operator and it's uh so um what do you think is changing for them for a tour operator in this new age of agentic ai i think it's a
Speaker 2big chance to finally solve the profitability problem because tour operators have always been very low margin um which partly comes out of that huge cost of operations and risk that they carry with them and one thing that we want to help them is actually to be able to do a lot of the work that they do and to be able to do a lot of the work that they do and to be able to do a lot of the work that they do and to be able to do a lot of the work actually make sure that it can be an economic and economically more stable and more profitable business in the future and then the other
Speaker 1one i want to get your take on and this comes back to how we started the conversation around the whole concept of thai workbench and moving from this idea of ai as a tool like a software tool um to actually being a digital workforce and that's when i was trying to kind of relate to our listeners what about significantly profound change and i always i always give um my friends and my colleagues a lot of attention to the when gilad um the gears over this because he would often say ai is just a tool which it felt very reassuring to a lot of travel advisors it was just a tool to accomplish your goals and as i've said to him and i've said on stage you know you look at someone like yuval noah harari which is one of my favorite authors and academics you know the author of sapiens and several other works and he's become a real expert in the ai space because um but he's a historian he's not like unlike you who has a understanding human behavior and how technology is going to impact human behavior and the point that he makes that um has always stayed with me is that never in the course of history have we ever created a tool that can think and act on our behalf and so it's not just a tool and so i'd love your view on this is like a lot of people have in the last few years man well been looking at ai as another software tool and they're using software tools that now have ai features so it being the best example with gemini that all of a sudden you're using ai mode and you're kind of like gradually transitioning over but when we think about agentic ai and tools acting on your behalf you prompting a platform to put an itinerary together to handle the full booking like actually like fully automating like what some people are doing with claude co-work now giving it complex tasks for it to perform on its own and then completing a board presentation for you in you know a matter of minutes that would have taken you five days previously so but tell us how you see this whole digital workforce because i'm sure some of the some of the companies listening to this because this is something i do see commonly and again why i wanted to ask you this question is there some there seems to be some hesitancy or reticence to embrace these tools because what it means for people inside the business today and um so they see i've seen teams marketing teams is a good example inside large travel insurance companies that have pushed back on using ai because they know what it's going to mean for replacing roles so it's like there seems a bit of a internal culture war in some companies where there's always a concern about change and what it means for us i mean as a species we are creatures of habit by nature and change is scary and it can be overwhelming so i wanted you to address that question and how best for our listeners to be thinking about that because if if you agree with my line of thinking um that it's not just a tool that this is now about having a digital workforce companies embracing nazaza are also going to have to look at their existing team and and rethink their workflows that needs to be part of how they're embracing this or at least that's the way i would see it but i'd love your view on
Speaker 2this yeah i mean about the digital workforce i think we're now at the stage of ai development and i mean nobody of us knows whether that's going to change with the super intelligence but then everything will be upside down anyway right but we're now at the stage where you can view an ai agent like an assistant right so you have somebody who does work for you still have to manage them you still have to give them instructions you have to check what they do you have to verify that it's correct so it's not a perfect implemented algorithm that always does the same perfectly but it's somebody who can do work on behalf of you and it takes a lot of time of you but you still have to watch out what they do you have to make sure that they do the right thing and you have to manage them and i think this this is the status today and yeah i think whether we like it or not it's the reality and so we have two choices either we embrace it or we don't and i actually i mean we we were rather early internally in doing this this ai transformation and making sure that our team embraces ai because um me and my leadership team we were absolutely convinced that there is no other way and i remember that time ago in an all hands what what i told my team was look the situation is the following there's a huge wave coming and we're not gonna stop this wave and now we have two choices either we go home or we learn to surf very well and very quickly and there is no other option right and so of course our choice was we have to learn to surf we have to learn how to ride the wave and how to use those and and then you can actually do things that before you were not able to do but if you're just sitting there and let and and watch the wave come it's going to wash over you and you're going to have a problem yeah and
Speaker 1that's why i was so excited to have this conversation with you because i see you very much ahead of the curve and ahead of the wave to use your expression um to be able to kind of see that wave coming and get up and actually ride that wave because that is a um to stay on that analogy like this this is a significant wave and it's going to be one that is going to um change so many aspects of our industry which is why i wanted to kind of take a step back and think big picture what this means where we're headed and obviously we're going to continue these conversations at our summit and i know many of our listeners are going to be keen just from this conversation uh to reach out to you and the team to learn more get a demo wrap their head around nazaza what you've built with um tie workbench where you're headed with tie guard um so let's make sure all of our listeners know how best to connect with you obviously i've talked directly manuel or the team what are the best ways to connect and partner with you guys
Speaker 2i mean you can always drop me or us an email um my email address is mh at nezaza with a set dot com and you can find me on linkedin um you can meet me or us at one of the upcoming travel conferences and shows so there are many ways reach out and i'm looking forward to have a conversation fantastic i just
Speaker 1want to encourage all of our listeners um to get a demo of this tech and obviously i think it's something that you would likely embrace tour operators out there, travel agencies. Obviously, it's interesting, even airlines. I know we've only just briefly touched on that. And that's why I'm excited for you to start to see other verticals. So even those of you that are not in those specific categories, I think it's actually worth understanding what Nazaz is building, where they're headed, and where that might be relevant for you and your business in the future. I mean, Manuel certainly has, you've got the sense on this podcast, a thought leader in this space. He's built a successful company. He's got a great team around him. And he's got some really compelling technology that is going to shape the way our travel industry functions in the future. So I just want to say thank you for all that you're doing for our industry, but also for making the time to join us and to be a part of our upcoming summit. And I certainly wish you and the team every success headed forward. And we'll look forward to keeping in touch with you as it
Speaker 2progresses. Thank you very much, Dan. Thank you very much, everybody who listens or will listen to this podcast. It was great. I had a great time and looking forward to all the future conversations.
Speaker 1Fantastic. Thank you, Manuel. And thank you to all of our listeners for taking the time to join Manuel and I to take a first look at Thai Workbench, which has just been announced. And we mentioned Thai Guard as well, which depending on when you're listening to this is likely live. If you haven't already registered for our Travel Trends Summit Beyond AI to see Manuel on October 29th. It's the 29th. As well, but he's speaking on the afternoon of October 28th. The full agenda is live on TravelTrendsPodcast.com and registration is free. So come and be part of the conversation. You can interact live and ask questions directly of Manuel as part of that session. And if you want to just dive right in, obviously, Nazaza.com, N-E-Z-A-S-A.com. You can find Manuel on LinkedIn for sure. Connect with his team. Check out the full team on LinkedIn. And M-H at Nazaza.com. Send Manuel a direct email. I'm sure he'd love feedback on the session. And just to keep in touch with those people in the industry that are intrigued to know what he's built with Nazaza and TripBuilder and what Thai Workbench could mean for the industry and where they're headed. So thanks again for joining us for this special spotlight episode. Don't forget, we do post clips and highlights on our social channels. You can find those on LinkedIn, Instagram, and YouTube. And make sure you're subscribed on the streaming platform of your choice. Whenever these new spotlight episodes go live. Thanks again for joining us. And until next time, safe travels.
PodcastTranscript AI